DEL RECONOCIMIENTO DE LOS HIJOS, NACIDOS FUERA DEL MATRIMONIO

A schematic diagram of the pendulum is shown in Figure 6.5. The pendulum is an aluminium structure that operates similarly to the Charpy impact test, an arm is pulled back to the release mechanism and released when required using a lever. Energy is lost as the pad scuffs the rail surface after the pendulum arm is released, higher friction levels result in a higher slip resistance value (SRV) which is read from the scale.

The pendulum has an arm, (c) on Figure 6.5, that pivots around point (b). A rubber pad (d), also known as the slider, is mounted to the end of the arm and shown in Figure 6.7. The slider can use various different pads but the hardest rubber has been chosen for this work as these have been used previously to obtain similar friction coefficients to those seen in hand pushed tribometer testing. The slider has 45 degree, 5 mm chamfer at the contact end and and is spring loaded to the pendulum arm which ensures that contact between pad and surface is maintained throughout the swing. The pendulum arm is swung from a set position which pushes the pointer up the scale, the arm and therefore the pointer travels further if there is less friction.

Figure 6.5: Schematic representation of the pendulum rig showing: (a) energy loss

scale; (b) height-adjustable pivot; (c) pendulum swing arm; (d) spring mounted slider (rubber pad attached here); (f ) height adjustable feet [27]

A wooden base has been manufactured to sit over a standard piece of rail so that the slider can contact, shown as used in the field in Figure 6.6. The wooden base is adjustable so that it can be used on uneven ground, the pendulum also has a bolt on each of its three feet that are used to keep the rig level, which can be checked using a built in spirit level.

Figure 6.6: Field testing on rail with the pendulum mount [27]

Figure 6.7: A pad slider with 5 mm chamfer

A 1000 mm section of 260 grade rail was used as the test surface. Surface rust was removed using an angle grinder and the rail head finished with p280 abrasive paper to produce a clean railhead contact patch that looked similar to those seen on track testing. The rail had been used previously so there was some surface pitting seen on the surface. A Four-S rubber pad with an IRHD hardness of 96 and was used for the majority of testing, as this was shown in previous tests to produce a consistent loss of force and the change in resistance was similar to those obtained from a hand pushed tribometer [65].

Another slider was manufactured with the same dimensions as the rubber cham- fered pad, but using EN24 steel so that a comparison could be made between the rubber/steel and a steel/steel contact. This had an identical chamfer to the rubber slider on one side and a rounded fillet on the other to use as a comparison.

6.2 Pendulum Testing

6.2.3 Methodology

The pad was conditioned between each set of tests and then cleaned between every test. The aim of conditioning was to make sure the rubber slider was clean and also had the same roughness and profile between each test. The pad was conditioned and returned to a standard finish by scuffing over P400 abrasive paper 3 times, followed by wetted 3M emery paper 10 times. The rail was cleaned between tests by scuffing over wetted 3M emery paper 3 times which removed any solid residue.

The pendulum was zeroed by adjusting the vertical adjustment knob and raising the pendulum arm to ensure there is no slider impact. The pendulum arm was released and caught on its return swing. The friction ring was adjusted in small increments using the clamp adjustment knob and the pendulum swung repeatedly until the pointer is pushed to zero for three consecutive swings.

The height of the pendulum arm was adjusted so that the spring loaded slider contacts the substrate for a distance of 127 mm during a swing. The pendulum was then swung using the following sequence:

1. The pendulum swing arm was raised and clipped in to its starting position. 2. The pendulum was released from its swing arm using a lever

3. Approaching 90 degrees into the swing, the rubber pad scuffed the test speci- men

4. The pointer was pushed along by the pendulum arm until it reached a maxi- mum value on the pendulum upstroke

5. The pendulum arm was caught by hand on the back swing to prevent any changes to the pointer

6. The energy lost due to friction was measured using the pointer position on a scale during the pendulum arm upstroke

The slip resistive value (SRV) can be converted to a comparable friction coefficient by using the equation shown in (6.1). It is important to remember that this is simply an approximation and not derived from the tangential and normal force as with the other rigs used in this work, but is useful when comparing approximate values. The high dry friction and low wet friction (approximately 0.6 and 0.15) do however provide compatible values to the dry and wet friction coefficients found on the other rigs [27]. For the purposes of comparison between other test method used in this work, pendulum results will be recorded as an approximate “friction coefficient”, converted from the SRV value using this equation [27].

µ = ( 110 SRV −

1 3)

−1 _(6.1)

The pad was conditioned and cleaned, then water was applied in different quantities to generate a basic understanding of how water will affect friction when using the pendulum rig to act as a baseline. Water was applied using a syringe, “drizzling” the

water over the largest area of railhead possible to maximise coverage. An average of 3 repeats were taken for each water volume, shown in Figure 6.3.

Another set of tests were carried out, applying water using a “spray bottle” rather than a syringe. This was a less accurate method of dispensing water but ensured better railhead coverage than a syringe. The spray bottle was found to dispense approximately 7 ml of water with each spray and the results for the average friction coefficient over 3 repeats for different numbers of sprays are shown in Table 6.4. The spray bottle lowered the friction coefficient more effectively with a smaller volume of water than when the same volume of water was applied using a syringe, which may be because a better surface coverage can be achieved with a spray bottle.

Water volume (ml) Average friction coefficient Standard deviation 0 0.72 0 2 0.35 0.0058 5 0.26 0.0058 10 0.21 0.012 20 0.18 0.012 30 0.18 0.012 40 0.18 0

Table 6.3: Friction coefficient dependence on volume of water, applied using a sy-

ringe, on a clean railhead

Number of sprays Estimated volume of water (ml) Average friction coefficient Standard deviation 0 0 0.74 0 1 7 0.19 0.0058 2 14 0.18 0.012 3 21 0.18 0.0058 4 28 0.18 0

Table 6.4: Friction coefficient dependence on volume of water, applied using a spray

bottle, on a clean railhead

6.2.3.1 Generated oxides

A method was created to test the changing friction of a drying, oxidised rail. Ap- plying a thin layer of water to the clean railhead and letting it evaporate was seen to produce a thin, orange layer of iron oxide. The oxide became visible approximately 100 seconds after water application and grew more visible as water evaporated, until

6.2 Pendulum Testing

the railhead had completely dried. A test method was established in order to com- pare the friction coefficient of this freshly oxidised railhead to a “cleaned” railhead. Two sets of tests were carried out, one using a pre-oxidised railhead and another using a railhead with no visible oxidation.

In the pre-oxidised set of tests the railhead was first cleaned using P400 abrasive paper to remove any oxide layer and then rubbed with a clean cloth soaked in acetone to remove any oil residue. The area of pendulum contact on the railhead was taped off with masking tape, so only the 127 mm long contact area was exposed. 1 spray of approximately 7 ml of water was applied to this contact area using a squirt bottle, creating a thin film of water. This water was left to dry which took approximately 8 minutes and this formed a thin orange film of iron oxide on the railhead. After 10 minutes the dry railhead was re-wetted with 1 spray of water using the spray bottle again and the pendulum swung every 20 seconds until the railhead was visibly dry and a dry slip resistance value was reached.

Another set of tests were carried out without the pre-oxidation step as a comparison, the railhead was cleaned using P400 abrasive paper and acetone, then the contact patch was once again taped off. The railhead was then wetted with 1 spray from the bottle but this time testing was started straight away so that the oxide layer was not formed. Once again the pendulum was swung every 20 seconds until a dry slip resistance value was reached. A summary of the two test methods is shown in Table 6.5.

Step Clean testing Pre-oxidised testing

1 Rail cleaned Rail cleaned 2 Rail wetted (7ml) Rail wetted (7ml) 3 Test until dry Wait 10 minutes

4 - Rail re-wetted

5 - Test until dry

Table 6.5: Test method to compare a clean and pre-oxidised railhead

6.2.3.2 Applied oxide

Magnetite and hematite oxide powder were also tested in paste form, as described in the introduction to this chapter, to assess any effects on friction levels. The pastes were made to a 50 wt % suspension using 3.5 g of oxide powder and 3.5 g of water. The 7 g of oxide paste was then spread over the surface of a freshly cleaned and non oxidised railhead, all the paste was added inside the area that the pendulum slider would contact and created a thin film of oxide paste that covered the contact patch. Tests were carried out using the same method used previously, the pastes were added and then a reading was taken every 20 seconds until the rail dried out and the friction coefficient returned to high values. As before, the results have been plotted against both time and number of pendulum swings.

A short series of tests under the same conditions were carried out with a pendulum slider that had been manufactured out of a piece of steel to assess if the steel on steel rather than rubber on steel contact would change any results.

6.2.3.3 Field testing

The effect of dew and weather conditions on the friction coefficient was assessed by testing on a piece of rail left outside so that the railhead could be exposed to natural conditions. The piece of rail was cleaned with abrasive paper and acetone in the evening and left overnight before readings were taken in the morning, beginning before dawn to assess if the dew point had a notable effect on friction characteristics. The rail was kept in an enclosed area outside the building, placed in between the outside building wall and a wooden fence which was surrounded by foliage. This means humidity levels were high and the rail was in a similar situation to that seen in a railway “cutting”.

The pendulum was zeroed and set up as outlined previously. The pendulum was left in the same position and not moved throughout the day, this meant that there was some wear of the rail surface oxide with each strike but it allowed comparable readings without having to set up the correct strike distance with every test which could cause further errors. This also better simulated the removal and regrowth of iron oxide by a wheel pass over a shorter time period rather than letting the railhead oxidise all day.

Three days were tested with readings at least every hour, but more regularly in the mornings as this was expected to be when dew was present and therefore important in the wet rail phenomenon.

The first two days contained a mixture of wet and dry conditions, on the first day the railhead surface oxide was abraded before the first pendulum reading to simulate a wheel pass and assess how quickly the friction coefficient would drop post abrasion. On the second day no morning abrasion was carried out so the railhead was oxidised heavily after being left outside the previous night.

The third morning of testing was used as a control, it was a dry day with a winds speed of 30 mph, which meant there was neither dew or precipitation and the rail- head remained dry.

6.2.4 Results

6.2.4.1 Generated oxide results

The “generated oxide” method was repeated 3 times each for both clean and pre- oxidised rail. Testing was performed at alternate times (clean, then pre-oxidised, then clean), over 2 days to help eliminate any issues with environmental changes

6.2 Pendulum Testing

such as temperature and humidity, along with physical changes such as pad wear or rail roughness. Average results are shown in Figure 6.8. The clean average friction coefficient starts at approximately 0.21 and drops slightly to 0.2 after 40 seconds, 2 swings. It then plateaus at 0.2 until 120 seconds and begins to climb until it reaches a dry value of 0.63 at 420 seconds. The friction coefficient then seems to drop slightly to 0.6 between 450 and 500 seconds.

The friction coefficient of the first swing of the pendulum using the pre-oxidised method averages slightly lower at 0.19. The friction coefficient then drops to ap- proximately 0.14 at 150 seconds, before rising slightly to 0.15 at 200 seconds. At this point the friction coefficient then climbs back up to a dry value similar to that seen in the clean testing at 430 seconds. An image of the railhead after oxidised testing is shown in Figure 6.9.

Figure 6.8: Average clean and pre-oxidised railhead friction coefficients

6.2.4.2 Applied oxide results

The applied magnetite and hematite oxide both followed a similar pattern, with average results for 3 repeats being shown in Figure 6.10. The initial pendulum swing would result in a fairly high friction coefficient, 0.26 for magnetite and 0.30 for hematite. On the second swing for both pastes, the friction coefficient was much lower, 0.21 for magnetite and 0.19 for hematite. At this point both friction coefficients levelled out before rising again on the fourth swing. Overall the hematite friction coefficient seemed to remain lower for longer, but neither produced a very low friction coefficient of below 0.1 as discussed in the literature review. The friction coefficient rose up to approximately 0.5 at 330 seconds for both sets of tests as the railhead dried out.

The previous pendulum tests, using dry and wet conditions, pre-oxidised rail and rail with applied oxide paste, were repeated with a steel slider. The results were plotted against the corresponding lowest friction coefficient for each condition using the rubber slider and shown in Figure 6.11. The greater mass and rigidity of the steel slider removed any oxide with a single strike so the readings for different conditions were very similar and did not vary as much as the rubber slider, the only large difference in friction coefficient was seen between wet and dry values with little difference between oxidised and non-oxidised conditions.

6.2 Pendulum Testing

Figure 6.11: Lowest friction coefficient recorded for different railhead conditions

using the steel and rubber pendulum sliders

6.2.4.3 Field testing results

Results for pendulum testing on a piece of rail outside in different conditions during November are shown in the following section. The friction coefficient was logged throughout the day, along with the relative humidity and whether there was precip- itation falling or dew on the railhead. Three days were tested, each with different weather conditions.

Day 1 began with dew on the rail at 7:00 am. There was then a period of dry weather before heavy rain at around 8:00 am. The railhead remained wet throughout the rest of the day because of light drizzle falling. Results are shown in 6.12.

On day 2, dew was again present in the morning, this time followed by a longer period of dry weather before the railhead was once again wetted by drizzle at 11:30. Results are shown in 6.13.

Finally, conditions over a dry and windy day in winter were plotted in 6.14, the railhead remained dry all day with no dew or precipitation.

Figure 6.12: Friction coefficient, weather conditions and relative humidity changes

6.2 Pendulum Testing

Figure 6.13: Friction coefficient, weather conditions and relative humidity changes

over day 2

Figure 6.14: Friction coefficient, weather conditions and relative humidity changes

6.2.5 Discussion

6.2.5.1 Generated oxide

The “generated” oxide laboratory testing showed a clear difference in the friction coefficient between a clean and pre-oxidised railhead. Compared to the clean rail- head, the pre-oxidised friction coefficient is lowered as soon as water is added and remains low after wetting for a period of time, before rapidly increasing as the rail dries. After prior oxidation the friction coefficient generally stays at a lower level (below 0.2) for a longer period of time than the rail that had not been pre-oxidised. The clean rail friction coefficient dips below 0.2 briefly, rising above it at 120 s. The friction coefficient of the oxidised rail on the other hand does not rise above 0.2 until 220 s have passed. The oxide layer is removed over a number of cycles and by the end of a 24 cycle test the layer has been virtually completely removed. The increase in friction coefficient seen during the pre-oxidised pendulum testing could be a combination of both the rail drying and the oxide being removed, the oxide seems to be removed at a quicker rate under dry conditions.

It is clear, from visual inspection, that although a thin layer of oxide is being formed in the wear track, the constant pendulum testing is suppressing the bulk of the oxide growth compared to that seen outside the wear track. It can also be noted that the rail used had large amounts of pitting. This creates a more patchy oxide layer which means that the friction coefficient may drop to lower levels if a less damaged railhead was used so a oxide layer with greater coverage could form. The pits would also likely produce a higher surface roughness, which as suggested in the modelling work may prevent low adhesion from occurring.

6.2.5.2 Applied oxides

The friction coefficients of both hematite and magnetite were above those seen in water so neither reduced the friction coefficient to a large extent in this set of tests. Hematite lowered the friction coefficient further than magnetite. The hematite was far more sticky so more remained on the railhead after the initial pendulum strike, whilst the magnetite seemed easier to wash off the railhead and when washed, did not leave as much debris behind.

The extra mass of the steel pendulum slider removed oxides in the wear area in a single swing and did not differentiate as much as the rubber slider between different conditions. Short, 3 swing tests were carried out when it became clear that the results would show little difference between conditions other than wet or dry. This highlights the difference between a sliding and a rolling/sliding contact, where any oxide worn away under a rolling/sliding test rig may remain in the contact, whilst any oxide under a purely sliding contact may simply be pushed aside so the third body layer cannot build up.

6.2 Pendulum Testing

The contact patch between rail and pendulum slider will be roughly the same size as the real wheel-rail contact patch but the pendulum is a purely sliding contact,